ICMCTF2005 Session TS1: The Atomistics of Thin Film Growth: Computational and Experimental Studies

The deposition of Pb atoms on Cu(111) produces two distinct surface phases: a random Pb-Cu surface-alloy phase and a Pb-overlayer phase. Within a specified range of Pb coverage and temperature, the two phases co-exist and spontaneously order into domain patterns with tens-of-nanometer periodicity [1]. The continuous evolution of the domain structures with increasing Pb coverage agrees well with theoretical descriptions based on competing long- and short-range interactions. The system thus provides an ideal model system for probing the interactions underlying two-dimensional self-assembly. Although a self-consistent picture of the thermodynamic driving forces responsible for pattern formation now exists a comprehensive understanding of the kinetic processes underlying self-assembly is only beginning to emerge. In this talk I will discuss recent low energy electron microscope/scanning tunneling microscope experiments and first-principles calculations that address the key question concerning the kinetics of pattern formation: how do individual domains of each phase, which contain 10s of thousands of atoms, assemble into ordered patterns in matters of seconds to minutes? I will show that for Pb/Cu(111) the self-assembly occurs by the collective motion of entire domains as a unit across the surface. I will present evidence that this motion is facile because (1) Pb atoms move quickly across the surface alloy due to a Pb-Cu exchange mechanism and (2) Cu atoms move quickly through the overlayer phase due to a high concentration of surface vacancies and mobile Cu atoms embedded within the overlayer. The latter finding is not at all intuitive and may explain why the remarkable pattern formation seen for Pb on Cu(111) has not been observed for other metal-metal systems.

*Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin company, for the U. S. Department of Energy under Contract #DE-AC04-94AL85000. [1] R. Plass, J. A. Last, N. C. Bartelt, and G. L. Kellogg, Nature 412, 875 (2001).

The work is aimed at the experimental and theoretical study of the radiation effect on the microstructure of niobium films being irradiated with their own ions during deposition.

Films were formed by the method of ion-atomic deposition. The essence of this method consists in film deposition from partially ionized metals vapor onto the substrate. Metal vapors were obtained by electron-beam heating of the material. The ion current density was controlled by the degree of vapor ionization in the low-voltage discharge of direct current. The ion energy was varied by applying the potential onto the substrate. The degree of ionization of the flow deposited was ranged from 0 to 0.45. The ion energy was from 5 to 500 eV.

The structure of niobium films was studied as a function of the degree of ionization of the flow deposited and of the ion energy. The X-ray diffraction methods were used to determine the changes in the mean-size block structure and microstrains of the crystalline lattice depending on the ion flow parameters. The measurements of the film density change with ion energy changing were carried out. The microstructure of films was examinated using the electron microscopy.

The computer simulation of the ion-atomic deposition of niobium films was performed under conditions corresponding to the experimentally observed ones. It is shown that the ion action exerts an influence on the growing film relief. An effect of ion densification is studied. The mechanisms of atomic ordering under ion irradiation are studied. It is shown that the radiation-induced defects have an influence on the process of film growth. The relation between the change in density and the size of the block film structure under ion densification has been revealed.

Nano-science and nano-technology have become the locus of attention of materials scientists in recent years. This has led to fruitful synergy between contemporaneous experimental and theoretical work. Two distinct examples from my own research where this has materialized will be shown.

1. Medium range order in amorphous silicon (a-Si): It has not been possible to measure atomic medium range order in disordered materials by conventional diffraction transmission electron microscopy experiments (TEM). A recently developed experimental technique of fluctuation TEM can be related to medium range order. Simulations for a family of atomistic structural models for unhydrogenated amorphous silicon suggest that fluctuation electron microscopy experiments have observed orientational order of para-crystalline grains in amorphous silicon. This order may consist of correlations in the orientation of nearby paracrystalline grains or anisotropy in the grain shape. This observation makes a natural connection to the known growth modes of microcrystalline silicon and may be useful for other materials systems. [Appl. Phys. Lett. 85, 745 (2004)].

2. Structure of Pt and RuPt clusters on graphite: Experimental results from deposition of Pt and Ru-Pt nano-clusters on carbon substrates show that (i) a cluster is formed of layers with fcc(111)-stacked cubo-octahedral geometry; (ii) Pt (in the Ru-Pt clusters) segregates to the top layer; and (iii) very small clusters (37 atoms) surprisingly show bulk-bond distances, despite their average co-ordination number being smaller than that in the bulk. Using density functional theory calculations, we explain these observations and provide a detailed characterization of the structure of the clusters. Surprisingly, we find that the graphitic carbon substrate causes metal-metal bonds in small clusters to expand to their bulk values despite the smaller surface lattice constant of graphite, contradicting a long-held belief that graphite is an inert growth template.